Remove line breaks

The mode you actually want

Nearly every line-break remover offers exactly one behaviour: delete every line break, everywhere. That is almost never what someone needs. It turns a three-page document into one unbroken paragraph, and the structure that made it readable is gone.

The common case is text copied out of a PDF, where a break appears at the end of every visual line rather than at the end of every sentence. What you want there is to rejoin the lines within each paragraph while keeping the blank line that separates one paragraph from the next. That is the default mode here.

Why PDFs do this

A PDF does not store paragraphs. It stores glyphs at coordinates on a page — this character at this position, that character next to it. Where a line ends is a consequence of the page layout, not of the writing.

When you select and copy, the viewer reconstructs a reading order and inserts a line break wherever the text moved down the page. The result looks correct in the PDF and arrives with a break every seventy characters. Nothing is broken; the paragraph structure simply was never stored in the first place.

Three ways to remove line breaks

Join paragraphs merges lines within each block of text and keeps blank lines between blocks. This is the PDF case, and the one you want most of the time.

Remove all collapses everything to a single line. Useful when pasting into a field that accepts only one line, or when preparing a value to drop into a single spreadsheet cell.

Trim trailing touches no line breaks at all. It removes spaces and tabs at the end of each line — invisible characters that quietly break Markdown formatting, produce phantom changes in version control, and make string comparisons fail for no visible reason.

Windows and Unix line endings

Text from Windows marks a line break with two characters, a carriage return followed by a newline. macOS and Linux use the newline alone. Mixing the two produces duplicate blank lines, and a stray carriage return is invisible in most editors while still breaking whatever reads the file next.

All three modes normalise line endings before doing anything else, so pasted Windows text behaves exactly like pasted Unix text and you never have to work out which one you have.